Actions35
- Genie Actions
- Databricks SQL Actions
- Unity Catalog Actions
- Model Serving Actions
- Files Actions
- Vector Search Actions
Overview
The node provides integration with Databricks API, specifically supporting multiple resources including "Vector Search". The "Scan Index" operation under the Vector Search resource allows users to scan or retrieve data from a specified vector search index within Databricks. This is useful for scenarios where you want to explore or fetch all entries in a vector index, such as for similarity searches, data analysis, or debugging your vector search setup.
Practical examples include:
- Scanning a vector index to retrieve all stored vectors for offline processing.
- Debugging or auditing the contents of a vector search index.
- Exporting vector data for use in other machine learning workflows.
Properties
| Name | Meaning |
|---|---|
| Index Name | Name of the vector search index |
This property is required and specifies which vector search index to scan.
Output
The output will contain a JSON field representing the scanned data from the specified vector search index. This typically includes the vector entries stored in that index. If the node supports binary data output (not explicitly shown here), it would represent any associated binary content related to the vector entries.
Dependencies
- Requires an active Databricks account with access to the Vector Search feature.
- Requires an API authentication token credential configured in n8n to authorize requests to the Databricks API.
- The node uses the base URL and authorization token from the provided credentials to communicate with Databricks.
Troubleshooting
- Common issue: Invalid or missing API token leading to authentication errors.
Resolution: Ensure the API token credential is correctly set up and has sufficient permissions. - Common issue: Specifying a non-existent or misspelled index name.
Resolution: Verify the exact name of the vector search index in Databricks before using it. - Error message: Authorization failures or 404 not found errors when scanning the index.
Resolution: Check network connectivity, credential validity, and that the index exists in the Databricks workspace.